The mammalian embryo and fetus are unable to develop without a well-established, functional placenta. This transitory yet indispensable structure attaches the conceptus to the uterus and establishes the vascular connections necessary for nutrient and gaseous exchange between maternal and fetal compartments. Genetic targeting strategy allows the generation of mice lacking a specific gene. Such approaches reveal: (i) the high incidence of mutant embryonic or fetal death in utero, and (ii) the extraembryonic (placental) causes of these deaths. Due to the similarities presented between mouse and human placenta, we propose to use the potential of mouse targeting experiments as a model in order to understand human obstetrical pathologies. In this paper, we first review genes that have been demonstrated to be required in mice for implantation, choriovitelline and chorioallantoic placentation. Using examples (integrins, homeoboxs, hepatocyte growth factor or epidermal growth factor receptor...) we demonstrate the reality and efficiency of such an approach. Other candidate genes (receptor of leukemia inhibitory factor, Wnt2 or retinoic acid receptor alpha...) in order to understand, prevent and treat human obstetrical pathologies.